Cyclohexanone (CAS 108-94-1, C6H10O, molar mass 98.15 g/mol) is recovered from cyclohexane oxidation KA oil or from phenol hydrogenation as a binary mixture with cyclohexanol (CAS 108-93-0, C6H12O, molar mass 100.16 g/mol). The normal boiling points are 155.6 °C and 161.1 °C at 101.3 kPa; the interval of 5.5 K places the pair in a difficult distillation envelope. Density at 20 °C is 0.9478 g/cm³ for cyclohexanone and 0.9624 g/cm³ for cyclohexanol. The Cyclohexanone-Cyclohexanol Azeotrope & Separation Data define the operating envelope for distillation-based purification of cyclohexanone. Published vapor-liquid equilibrium data for the binary show a low relative volatility, typically 1.08–1.35 in the high-cyclohexanone region, and some datasets contain a minimum-boiling azeotrope or tangent pinch at cyclohexanone mole fractions above 0.95. The azeotropic composition is pressure-sensitive and is central to any separation data package for cyclohexanone finishing.

What are the reported azeotropic coordinates for the binary at 101.3 kPa?

At 101.3 kPa, the closest agreement among published isobaric VLE correlations places the azeotrope at approximately 154.6 °C–155.2 °C with a cyclohexanone liquid mole fraction of 0.95–0.98; vapor and liquid compositions are equal at the azeotrope. The scatter arises from differences in sample dehydration, gas chromatograph calibration, pressure control, and the tendency of cyclohexanol to solidify at 25.2 °C. Transfer lines, receivers, and VLE cells should be maintained above 30 °C to prevent partial condensation that biases x,y data.

Recirculating stills of the Gillespie type and dynamic ebulliometers are preferred for this binary because the low relative volatility demands x,y data accurate to better than 0.0005 mole fraction and 0.05 °C in bubble-point temperature. Vapor-pressure measurements may be screened under ASTM E1719 and ASTM E1194 procedures; however, those methods alone do not provide the liquid-phase composition needed for azeotrope assignment. Water must be excluded to ≤0.01 wt% because cyclohexanone forms a water heteroazeotrope at approximately 95 °C, shifting both the bubble point and the apparent relative volatility.

Vacuum operation shifts the azeotropic composition when the minimum-boiling condition persists. For a minimum-boiling azeotrope, the azeotropic mole fraction generally moves toward the component with higher heat of vaporization as pressure is reduced; for cyclohexanone-cyclohexanol, published pressure-scan data are limited, and an engineering data package should require a binary VLE regression using NRTL or UNIQUAC activity-coefficient models fitted to isobaric data at the planned column pressure. Use of a single atmospheric data point without pressure correction is not reliable below 20 kPa.

Activity-coefficient models for the binary require at least three isobaric datasets covering the column pressure range. NRTL regressions typically use a fixed nonrandomness parameter of 0.3. The binary is also sensitive to the Antoine constants selected for vapor pressure. An error of 0.5% in vapor pressure at 50 kPa can shift the predicted azeotropic composition by 0.01 mole fraction. Simulation files should therefore be validated against bubble-point data at the actual column overhead and reboiler pressures.

Industrial separation trains combine staged rectification with solvent-based or membrane steps

Cyclohexanone product at ≥99.5 wt% can be obtained from cyclohexanone-cyclohexanol feeds by high-theoretical-stage vacuum distillation only if the column is designed for the low relative volatility. Structured gauze packing in production-scale columns has been specified for 60–80 theoretical stages at reflux ratios of 15:1–30:1 for separation in the 0.10–0.90 mole fraction cyclohexanone range. Pressure drop per theoretical stage should remain below 0.3 kPa; reboiler wall temperatures above 180 °C accelerate aldol condensation and discoloration, so vacuum operation at 10–50 kPa is preferred. Falling-film or thin-film reboilers are used in place of kettle reboilers to reduce residence time at high temperature.

The overhead condenser and reflux drum are designed for cyclohexanone-rich liquid at 40–80 °C under vacuum. Because cyclohexanone has a flash point of 44 °C, hot lines and the reflux drum should be inerted with nitrogen to keep oxygen below 8 vol%. A vent condenser is installed downstream to recover cyclohexanone from the vacuum ejector load. Liquid hold-up in the reflux drum is limited to 3–5 min because prolonged contact with a recycled aqueous phase can promote cyclohexanol back-formation.

Diethylene glycol and ethylene glycol are used as extractive distillation solvents because hydroxyl groups associate preferentially with cyclohexanol, increasing the vapor-phase fraction of cyclohexanone. A solvent/feed mass ratio of 0.5:1–2:1 is typical. The entrainer recovery column is operated at 20–40 kPa to keep the reboiler temperature below 120 °C; higher temperatures cause glycol decomposition and acid formation that can catalyze cyclohexanone aldolization.

The solvent recovery column separates cyclohexanol from diethylene glycol. Cyclohexanol is removed overhead at 10–20 kPa and is subsequently dehydrogenated to cyclohexanone over a copper-zinc catalyst at 220–260 °C in a separate reactor. The lean glycol is cooled to 40 °C before returning to the extractive column; thermal degradation is monitored by acid number, and spent glycol is purged at 2–5 wt% of circulation to prevent accumulation of high-boiling esters.

If high-purity cyclohexanone is required beyond the azeotrope, extractive distillation becomes the controlling step

When the atmospheric azeotrope is accepted at 0.95–0.98 mole fraction cyclohexanone, simple rectification cannot reach ≥99.9 wt% cyclohexanone from an azeotropic feed. Extractive distillation with a polar glycol changes the activity coefficient of cyclohexanol more strongly than that of cyclohexanone. Process simulations using NRTL or UNIQUAC models fitted to binary VLE data indicate that the relative volatility increases from 1.1 to 1.8–2.0 when the entrainer concentration in the liquid phase exceeds 0.45 mass fraction. This allows the cyclohexanone overhead to exceed 99.5 wt% without a pressure-swing requirement. A two-column extractive train consists of a main extractive distillation column, a solvent recovery column, and a solvent cooler. The main column must have 30–45 theoretical stages, with feed and solvent introduced at separate points; solvent feed is preheated to 80–90 °C to prevent viscosity-related maldistribution.

Pressure-swing distillation is sometimes considered but is not normally used for cyclohexanone-cyclohexanol because the azeotropic composition shifts only weakly within the operable pressure range and the pair does not form a pressure-insensitive heterogeneous azeotrope. A two-column pressure-swing design would require operating pressures near 5 kPa and 800 kPa, with the high-pressure column exposing cyclohexanone to reboiler temperatures above 180 °C. This is generally unattractive for a thermally sensitive ketone.

Membrane-based vapor permeation with hydrophobic silicalite-1 or polyvinyl alcohol membranes has been proposed for ketone-alcohol separation. Pilot results show cyclohexanone/cyclohexanol selectivity of 2–4 at 60–80 °C, but published data for this specific configuration are limited, and membrane stability in the presence of high-concentration cyclohexanol is not fully documented. The technology is not considered a replacement for extractive distillation in current production-scale KA oil finishing.

Cyclohexanone storage should exclude oxygen and direct sunlight because autoxidation can generate cyclohexanone peroxides. Peroxide formation is controlled by nitrogen blanketing rather than by routine addition of antioxidants unless supplier validation supports a specific stabilizer. Avoid contact with strong oxidizing agents, strong acids, and amines because aldol condensation and Schiff-base formation can alter the distillation behavior and foul reboiler surfaces.

Analytical verification and control of the azeotrope-bounded overhead

Offline gas chromatographic analysis of cyclohexanone and cyclohexanol is carried out on a polar fused-silica column with thermal conductivity detection. Because no single ASTM method is specific to this binary, the GC method is calibrated against certified reference materials and verified by ASTM D4052-18a density measurements at 20 °C. The density of cyclohexanone is 0.9478 g/cm³ and that of cyclohexanol is 0.9624 g/cm³; this difference supports composition cross-checks, but the method is not sufficiently sensitive for azeotrope-bounded control by itself.

Karl Fischer water titration should be completed before VLE evaluation because water forms a cyclohexanone-water heteroazeotrope at approximately 95 °C; at water mass fractions above 0.1 wt%, the apparent azeotrope shifts and binary VLE data are not valid. Sample loops are maintained at 35 °C and purged with dry nitrogen to prevent cyclohexanol solidification and carbonyl hydration.

Process control on the finishing column uses online gas chromatography with a sampling interval of 3–10 min. The control range between the cyclohexanone-cyclohexanol azeotrope and the product specification is narrower than 0.5 wt% when the atmospheric azeotrope is accepted at 0.95–0.98 mole fraction. Feed-composition disturbances greater than 2 wt% require an immediate change in draw-off ratio or entrainer addition; otherwise the overhead slips into the azeotropic composition and off-spec product is produced.